Abstract 4137243: Dynamic changes of pericoronary fat attenuation index predicting changes of coronary plaque burden

Background: Percent atheroma volume (PAV) is a validated index of coronary plaque burden, linking to adverse cardiac events. Meanwhile, pericoronary fat attenuation index (FAI) on coronary computed tomography angiography (CCTA) image is a novel prognostic marker of coronary inflammation. Research Questions: It is unknown the impact of coronary inflammation (i.e., FAI) on the progression of coronary PAV in patients at an increased risk of adverse cardiac events. Methods: Patients with type 2 diabetes mellitus (T2DM) who underwent at least two clinically indicated CCTA examinations (time interval > 1 year) at our center were potentially eligible for enrollment. Eligible patients presented with more than one study plaque, defined as non-obstructive stenosis at baseline and not being intervened during serial CCTA. Longitudinal CCTA images were assessed for FAI and compositional plaque characteristics (e.g., PAV) using dedicated post-processing softwares. The relationships between changes in FAI and progression of compositional PAV were examined using the logistic and linear regression analysis. Results: A total of 204 T2DM patients (mean age, 61.2 ± 9.3 years; male, 68.1%) with 366 study plaques were enrolled. All patients were treated with statins, and their time interval between longitudinal CCTAs was 14.1 (12.9, 17.6) months. Compared with patients with improved coronary inflammation (ΔFAI ≤ 0), those with deteriorated coronary inflammation (ΔFAI > 0) had lower high density lipoprotein cholesterol and baseline FAI, whereas higher blood pressure (all p < 0.05). Meanwhile, compared to patients with ΔFAI ≤ 0, those with ΔFAI > 0 had more Δoverall and Δnon-calcified PAV (all p < 0.05). Importantly, these findings were consistent in a dedicated propensity score matching analysis. In logistic regression analysis, ΔFAI > 0 was associated with annual progression of overall PAV and non-calcified PAV, whereas ΔFAI ≤ 0 associated with annual regression of overall PAV and non-calcified PAV after adjusting for cardiac risk factors, baseline PAV and medications (all p < 0.05). Similarly, in linear regression analysis, changes of FAI were independently positively associated with annual changes of overall PAV (β=0.545, p< 0.001) and non-calcified PAV (β=0.461, p< 0.001) (Figure 1). Conclusion: In this longitudinal CCTA study, changes of coronary inflammation are closely correlated with changes of overall and non-calcified PAV among T2DM patients treated with statins.